ABSTRACTCancers rely on multiple, heterogeneous processes at different scales, pertaining to many... more ABSTRACTCancers rely on multiple, heterogeneous processes at different scales, pertaining to many biomedical fields. Therefore, understanding cancer is necessarily an interdisciplinary task that requires placing specialised experimental and clinical research into a broader conceptual, theoretical, and methodological framework. Without such a framework, oncology will collect piecemeal results, with scant dialogue between the different scientific communities studying cancer. We argue that one important way forward in service of a more successful dialogue is through greater integration of applied sciences (experimental and clinical) with conceptual and theoretical approaches, informed by philosophical methods. By way of illustration, we explore six central themes: (i) the role of mutations in cancer; (ii) the clonal evolution of cancer cells; (iii) the relationship between cancer and multicellularity; (iv) the tumour microenvironment; (v) the immune system; and (vi) stem cells. In each...
Background: Metastasis is the leading cause of cancer related deaths in breast cancer patients. L... more Background: Metastasis is the leading cause of cancer related deaths in breast cancer patients. Lymphovascular invasion represents one of the earliest stages of metastasis wherein the cells are introduced to a very different and distinct microenvironment. Methods: We leveraged spatial techniques developed for limited specimens in archival tissue to study patient matched cross-sectional tumor samples from different stages of breast neoplasia including normal breast, ductal carcinoma in situ (DCIS), primary invasive carcinoma (IBC), lymphovascular invasion (LVI) and regional lymph node metastasis. We selected a set of 21 patients with ER+ breast cancer to generate cross-sectional samples of each of these stages, for a total of 331 samples. The areas of LVI were identified by a combination of H&E review and immunohistochemistry for podoplanin. We performed smart-3SEQ for gene expression profiling and light pass whole genome sequencing for DNA copy number alterations. Results: We profil...
Hypoxia promotes aggressive tumor phenotypes and mediates the recruitment of suppressive T cells ... more Hypoxia promotes aggressive tumor phenotypes and mediates the recruitment of suppressive T cells in invasive breast carcinomas. We investigated the role of hypoxia in relation to T-cell regulation in ductal carcinoma in situ (DCIS). We designed a deep learning system tailored for the tissue architecture complexity of DCIS, and compared pure DCIS cases with the synchronous DCIS and invasive components within invasive ductal carcinoma cases. Single-cell classification was applied in tandem with a new method for DCIS ductal segmentation in dual-stained CA9 and FOXP3, whole-tumor section digital pathology images. Pure DCIS typically has an intermediate level of colocalization of FOXP3+ and CA9+ cells, but in invasive carcinoma cases, the FOXP3+ (T-regulatory) cells may have relocated from the DCIS and into the invasive parts of the tumor, leading to high levels of colocalization in the invasive parts but low levels in the synchronous DCIS component. This may be due to invasive, hypoxic ...
Understanding intratumor heterogeneity is critical for studying tumorigenesis and designing perso... more Understanding intratumor heterogeneity is critical for studying tumorigenesis and designing personalized treatments. To decompose the mixed cell population in a tumor, subclones are inferred computationally based on variant allele frequency (VAF) from bulk sequencing data. In this study, we showed that sequencing depth, mean VAF, and variance of VAF of a subclone are confounded. Without considering this effect, current methods require deep-sequencing data (>300× depth) to reliably infer subclones. Here, we present a novel algorithm that incorporates depth-variance and mean-variance dependencies in a clustering error model and successfully identifies subclones in tumors sequenced at depths of as low as 30×. We implemented the algorithm as a model-based adaptive grouping of subclones (MAGOS) method. Analyses of computer simulated data and empirical sequencing data showed that MAGOS outperformed existing methods on minimum sequencing depth, decomposition accuracy, and computation ef...
Introduction. As nonobligate precursors of invasive disease, pre-cancers provide a unique vantage... more Introduction. As nonobligate precursors of invasive disease, pre-cancers provide a unique vantage point from which to study the molecular pathways and evolutionary dynamics that lead to the development of life-threatening cancers. Ductal carcinoma in situ (DCIS) is the most commonly diagnosed precursor of breast cancer, with variable propensity for invasive progression. In order to address the problems of over- and under-treatment, we performed a multimodal, integrated profile of DCIS with clinical outcomes with which to develop and validate predictors of invasive progression. Methods. We present observations on DNA, RNA, and protein expression on two independent patient cohorts of DCIS, diagnosed from 1981 to 2014, from the Translational Breast Cancer Research Consortium (TBCRC 038) and the Washington University Repository of Archival Human Breast Tissue (RAHBT). Patients initially diagnosed with DCIS, with either DCIS or invasive recurrence (cases; mean follow up 5.8 years) were m...
2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI)
Ductal Carcinoma In Situ (DCIS) is a non-obligatory precursor of Invasive Breast Cancer. It is th... more Ductal Carcinoma In Situ (DCIS) is a non-obligatory precursor of Invasive Breast Cancer. It is the most common mammographically detected breast cancer. Predicting DCIS progression to invasive ductal carcinoma is a major clinical challenge due to the lack of a uniform classification system in the diagnosis and prognostication of this disease. To characterise the tissue microecology of DCIS, we proposed and tested the model "DCIS-Identification model" based on Generative Adversarial Networks (GAN) for detection and segmentation of DCIS ducts from multiplex immunohistochemistry (IHC) staining samples. We also trained a Spatially Constrained Convolutional Neural Network (SC-CNN) to detect and classify single cells based on their CA9 and FOXP3 expression. The DCIS-Identification model was evaluated on 8 whole slide images, resulting in an average Dice score of 0.95 for the segmentation performance. The single cell identification framework was tested on 10 randomly selected whole slide sections, achieving the average accuracy of 88.6% in a 5 fold cross validation scheme. With the proposed pipeline, we efficiently integrated deep learning, computational pathology and spatial statistics to report distinct differences in the microenvironments of DCIS and IDC/DCIS samples. The proposed pipeline provides a tool for a better understanding of the mechanism of tumours in DCIS and IDC/DCIS cases.
<b>Summary</b> This metadata record provides details of the data supporting the claim... more <b>Summary</b> This metadata record provides details of the data supporting the claims of the related manuscript: "Unmasking the immune microecology of ductal carcinoma in situ with deep learning". The data consist of immunohistochemistry (IHC) haematoxylin and eosin (H&amp;E) staining images of grade 2-3 pure ductal carcinoma in situ (DCIS) and DCIS adjacent to invasive cancer (adjacent DCIS) samples. The related study aimed to characterise tissue spatial architecture and the microenvironment of DCIS via design and validation of a new deep learning pipeline. <b> </b> <b>Data access</b> All training data, including the fully anonymised raw H&amp;E image tiles and pathological annotations as binary marks, as well as Python code, are available in the corresponding author's GitHub: https://github.com/pathdata/HE_Tissue_Segmentation. Requests for data access for the Duke samples can be submitted to E. Shelley Hwang ([email protected])<sup> </sup>and Yinyin Yuan ([email protected]). Data underlying Figures 4 and 6 are in the files "Ext_validData_DCIS_DAVE_Fig4_data.csv" and "Ext_validData_DCIS_DAVE_Fig6_data.csv", included with this metadata record. The images used as representative examples in Figure 8 are listed in the file "Figure 8 image details.xlsx", included with this metadata record. <b>Name of Institutional Review Board or ethics committee that approved the study</b> The study was approved by the institutional review board of Duke with a waiver of the requirement to obtain informed consent.
SUMMARYDuctal carcinoma in situ (DCIS) is the most common precursor of invasive breast cancer (IB... more SUMMARYDuctal carcinoma in situ (DCIS) is the most common precursor of invasive breast cancer (IBC), with variable propensity for progression. We have performed the first multiscale, integrated profiling of DCIS with clinical outcomes by analyzing 677 DCIS samples from 481 patients with 7.1 years median follow-up from the Translational Breast Cancer Research Consortium (TBCRC) 038 study and the Resource of Archival Breast Tissue (RAHBT) cohorts. We identified 812 genes associated with ipsilateral recurrence within 5 years from treatment and developed a classifier that was predictive of DCIS or IBC recurrence in both cohorts. Pathways associated with recurrence include proliferation, immune response, and metabolism. Distinct stromal expression patterns and immune cell compositions were identified. Our multiscale approach employed in situ methods to generate a spatially resolved atlas of breast precancers, where complementary modalities can be directly compared and correlated with con...
Supplementary Figures and Tables. This file contains the following figures and tables: Validation... more Supplementary Figures and Tables. This file contains the following figures and tables: Validation of image analysis with pathological scores. Figure S2. Comparison of Voronoi and Square tessellation. Figure S3. Comparing the Morisita-Horn index with Pearson correlation. Figure S4. Heatmap to show correlation among Pearson correlation and the Morisita-Horn index computed over different scales with square (S) or Voronoi (V) tessellation. Figure S5. Association of immune-cancer correlation with survival according to different spatial scales. Figure S6. Optimizing the Morisita index cutoff in the discovery cohort. Figure S7. Immune-cancer colocalization sub-stratifies clinical parameters and immune abundance, estrogen receptor (ER) status and TP53 mutation status in human epidermal growth factor receptor 2 (HER2)â +â subtype. Figure S8. Robustness of prognostic value of Morisita index in HER2-amplified subtype. Figure S9. Intra-slide score variability. Figure S10. Kaplan-Meier curves of...
In mammography, calcifications are one of the most common signs of breast cancer. Detection of su... more In mammography, calcifications are one of the most common signs of breast cancer. Detection of such lesions is an active area of research for computer-aided diagnosis and machine learning algorithms. Due to limited numbers of positive cases, many supervised detection models suffer from overfitting and fail to generalize. We present a one-class, semi-supervised framework using a deep convolutional autoencoder trained with over 50,000 images from 11,000 negative-only cases. Since the model learned from only normal breast parenchymal features, calcifications produced large signals when comparing the residuals between input and reconstruction output images. As a key advancement, a structural dissimilarity index was used to suppress non-structural noises. Our selected model achieved pixel-based AUROC of 0.959 and AUPRC of 0.676 during validation, where calcification masks were defined in a semi-automated process. Although not trained directly on any cancers, detection performance of calcification lesions on 1,883 testing images (645 malignant and 1238 negative) achieved 75% sensitivity at 2.5 false positives per image. Performance plateaued early when trained with only a fraction of the cases, and greater model complexity or a larger dataset did not improve performance. This study demonstrates the potential of this anomaly detection approach to detect mammographic calcifications in a semi-supervised manner with efficient use of a small number of labeled images, and may facilitate new clinical applications such as computer-aided triage and quality improvement.
Beneficial mutations that arise in an evolving asexual population may compete or interact in ways... more Beneficial mutations that arise in an evolving asexual population may compete or interact in ways that alter the overall rate of adaptation through mechanisms such as clonal or functional interference. The application of multiple selective pressures simultaneously may allow for a greater number of adaptive mutations, increasing the opportunities for competition between selectively advantageous alterations, and thereby reducing the rate of adaptation. We evolved a strain of Saccharomyces cerevisiae that could not produce its own histidine or uracil for ~500 generations under one or three selective pressures: limitation of the concentration of glucose, histidine, and/or uracil in the media. The rate of adaptation was obtained by measuring evolved relative fitness using competition assays. Populations evolved under a single selective pressure showed a statistically significant increase in fitness on those pressures relative to the ancestral strain, but the populations evolved on all th...
Crucial transitions in cancer-including tumor initiation, local expansion, metastasis, and therap... more Crucial transitions in cancer-including tumor initiation, local expansion, metastasis, and therapeutic resistance-involve complex interactions between cells within the dynamic tumor ecosystem. Transformative single-cell genomics technologies and spatial multiplex in situ methods now provide an opportunity to interrogate this complexity at unprecedented resolution. The Human Tumor Atlas Network (HTAN), part of the National Cancer Institute (NCI) Cancer Moonshot Initiative, will establish a clinical, experimental, computational, and organizational framework to generate informative and accessible three-dimensional atlases of cancer transitions for a diverse set of tumor types. This effort complements both ongoing efforts to map healthy organs and previous largescale cancer genomics approaches focused on bulk sequencing at a single point in time. Generating single-cell, multiparametric, longitudinal atlases and integrating them with clinical outcomes should help identify novel predictive biomarkers and features as well as therapeutically relevant cell types, cell states, and cellular interactions across transitions. The resulting tumor atlases should have a profound impact on our understanding of cancer biology and have the potential to improve cancer detection, prevention, and therapeutic discovery for better precision-medicine treatments of cancer patients and those at risk for cancer. Cancer forms and progresses through a series of critical transitions-from pre-malignant to malignant states, from locally contained to metastatic disease, and from treatment-responsive to treatment-resistant tumors (Figure ). Although specifics differ across tumor types and patients, all transitions involve complex dynamic interactions between diverse pre-malignant, malignant, and non-malignant cells (e.g., stroma cells and immune cells), often organized in specific patterns within the tumor
Despite increasing evidence supporting the clinical relevance of tumour infiltrating lymphocytes ... more Despite increasing evidence supporting the clinical relevance of tumour infiltrating lymphocytes (TILs) in invasive breast cancer, TIL spatial distribution pattern surrounding ductal carcinoma in situ (DCIS) and its association with progression is not well understood.To characterize the tissue microecology of DCIS, we designed and tested a new deep learning pipeline, UNMaSk (UNet-IM-Net-SCCNN), for the automated detection and simultaneous segmentation of DCIS ducts. This new method achieved the highest sensitivity and recall over cutting-edge deep learning networks in three patient cohorts, as well as the highest concordance with DCIS identification based on CK5 staining.Following automated DCIS detection, spatial tessellation centred at each DCIS duct created the boundary in which local ecology can be studied. Single cell identification and classification was performed with an existing deep learning method to map the distribution of TILs. In a dataset comprising grade 2-3 pure DCIS...
Understanding intratumor heterogeneity is critical to designing personalized treatments and impro... more Understanding intratumor heterogeneity is critical to designing personalized treatments and improving clinical outcomes of cancers. Such investigations require accurate delineation of the subclonal composition of a tumor, which to date can only be reliably inferred from deep-sequencing data (>300x depth). To enable accurate subclonal discovery in tumors sequenced at standard depths (30-50x), we develop a novel computational method that incorporates an adaptive error model into statistical decomposition of mixed populations, which corrects the mean-variance dependency of sequencing data at the subclonal level. Tested on extensive computer simulations and real-world data, this new method, named model-based adaptive grouping of subclones (MAGOS), consistently outperforms existing methods on minimum sequencing depth, decomposition accuracy and computation efficiency. MAGOS supports subclone analysis using single nucleotide variants and copy number variants from one or more samples of...
Kombucha, a fermented tea beverage with an acidic and effervescent taste, is composed of a multis... more Kombucha, a fermented tea beverage with an acidic and effervescent taste, is composed of a multispecies microbial ecosystem with complex interactions that are characterized by both cooperation and conflict. In kombucha, a complex community of bacteria and yeast initiates the fermentation of a starter tea (usually black or green tea with sugar), producing a biofilm that covers the liquid over several weeks. This happens through several fermentative phases that are characterized by cooperation and competition among the microbes within the kombucha solution. Yeast produce invertase as a public good that enables both yeast and bacteria to metabolize sugars. Bacteria produce a surface biofilm which may act as a public good providing protection from invaders, storage for resources, and greater access to oxygen for microbes embedded within it. The ethanol and acid produced during the fermentative process (by yeast and bacteria, respectively) may also help to protect the system from invasio...
Kombucha is a sweetened tea fermented by bacteria and yeast into a carbonated, acidic drink, prod... more Kombucha is a sweetened tea fermented by bacteria and yeast into a carbonated, acidic drink, producing a surface biofilm pellicle (colloquially called a SCOBY) during the process. Typically, liquid and a biofilm pellicle from a previously fermented culture is used as a starter for new cultures; however, there is no standard protocol for growing kombucha in the laboratory. In order to establish a standard protocol with low variability between replicates, we tested whether we could begin a kombucha culture with only well-mixed liquid stock. We found that viable kombucha cultures can be grown from low percentages of initial inoculum stock liquid, that new pellicles can form from liquid alone (with no ‘starter’ pellicle), and that the variation in the pellicle characteristics is lower when only a liquid starter is used (p = 0.0004). We also found that blending the pellicle before including it significantly reduces the variation among replicates, though the final pellicle was abnormal. W...
Within the cancer microenvironment, the growth and proliferation of cancer cells in the primary s... more Within the cancer microenvironment, the growth and proliferation of cancer cells in the primary site as well as in the metastatic site represent a global biological phenomenon. To understand the growth, proliferation and progression of cancer either by local expansion and/or metastasis, it is important to understand the cancer microenvironment and host response to cancer growth. Melanoma is an excellent model to study the interaction of cancer initiation and growth in relationship to its microenvironment. Social evolution with cooperative cellular groups within an organism is what gives rise to multicellularity in the first place. Cancer cells evolve to exploit their cellular environment. The foundations of multicellular cooperation break down in cancer because those cells that misbehave have an evolutionary advantage over their normally behaving neighbors. It is important to classify evolutionary and ecological aspects of cancer growth, thus, data for cancer growth and outcomes need to be collected to define these parameters so that accurate predictions of how cancer cells may proliferate and metastasize can be developed.
ABSTRACTCancers rely on multiple, heterogeneous processes at different scales, pertaining to many... more ABSTRACTCancers rely on multiple, heterogeneous processes at different scales, pertaining to many biomedical fields. Therefore, understanding cancer is necessarily an interdisciplinary task that requires placing specialised experimental and clinical research into a broader conceptual, theoretical, and methodological framework. Without such a framework, oncology will collect piecemeal results, with scant dialogue between the different scientific communities studying cancer. We argue that one important way forward in service of a more successful dialogue is through greater integration of applied sciences (experimental and clinical) with conceptual and theoretical approaches, informed by philosophical methods. By way of illustration, we explore six central themes: (i) the role of mutations in cancer; (ii) the clonal evolution of cancer cells; (iii) the relationship between cancer and multicellularity; (iv) the tumour microenvironment; (v) the immune system; and (vi) stem cells. In each...
Background: Metastasis is the leading cause of cancer related deaths in breast cancer patients. L... more Background: Metastasis is the leading cause of cancer related deaths in breast cancer patients. Lymphovascular invasion represents one of the earliest stages of metastasis wherein the cells are introduced to a very different and distinct microenvironment. Methods: We leveraged spatial techniques developed for limited specimens in archival tissue to study patient matched cross-sectional tumor samples from different stages of breast neoplasia including normal breast, ductal carcinoma in situ (DCIS), primary invasive carcinoma (IBC), lymphovascular invasion (LVI) and regional lymph node metastasis. We selected a set of 21 patients with ER+ breast cancer to generate cross-sectional samples of each of these stages, for a total of 331 samples. The areas of LVI were identified by a combination of H&E review and immunohistochemistry for podoplanin. We performed smart-3SEQ for gene expression profiling and light pass whole genome sequencing for DNA copy number alterations. Results: We profil...
Hypoxia promotes aggressive tumor phenotypes and mediates the recruitment of suppressive T cells ... more Hypoxia promotes aggressive tumor phenotypes and mediates the recruitment of suppressive T cells in invasive breast carcinomas. We investigated the role of hypoxia in relation to T-cell regulation in ductal carcinoma in situ (DCIS). We designed a deep learning system tailored for the tissue architecture complexity of DCIS, and compared pure DCIS cases with the synchronous DCIS and invasive components within invasive ductal carcinoma cases. Single-cell classification was applied in tandem with a new method for DCIS ductal segmentation in dual-stained CA9 and FOXP3, whole-tumor section digital pathology images. Pure DCIS typically has an intermediate level of colocalization of FOXP3+ and CA9+ cells, but in invasive carcinoma cases, the FOXP3+ (T-regulatory) cells may have relocated from the DCIS and into the invasive parts of the tumor, leading to high levels of colocalization in the invasive parts but low levels in the synchronous DCIS component. This may be due to invasive, hypoxic ...
Understanding intratumor heterogeneity is critical for studying tumorigenesis and designing perso... more Understanding intratumor heterogeneity is critical for studying tumorigenesis and designing personalized treatments. To decompose the mixed cell population in a tumor, subclones are inferred computationally based on variant allele frequency (VAF) from bulk sequencing data. In this study, we showed that sequencing depth, mean VAF, and variance of VAF of a subclone are confounded. Without considering this effect, current methods require deep-sequencing data (>300× depth) to reliably infer subclones. Here, we present a novel algorithm that incorporates depth-variance and mean-variance dependencies in a clustering error model and successfully identifies subclones in tumors sequenced at depths of as low as 30×. We implemented the algorithm as a model-based adaptive grouping of subclones (MAGOS) method. Analyses of computer simulated data and empirical sequencing data showed that MAGOS outperformed existing methods on minimum sequencing depth, decomposition accuracy, and computation ef...
Introduction. As nonobligate precursors of invasive disease, pre-cancers provide a unique vantage... more Introduction. As nonobligate precursors of invasive disease, pre-cancers provide a unique vantage point from which to study the molecular pathways and evolutionary dynamics that lead to the development of life-threatening cancers. Ductal carcinoma in situ (DCIS) is the most commonly diagnosed precursor of breast cancer, with variable propensity for invasive progression. In order to address the problems of over- and under-treatment, we performed a multimodal, integrated profile of DCIS with clinical outcomes with which to develop and validate predictors of invasive progression. Methods. We present observations on DNA, RNA, and protein expression on two independent patient cohorts of DCIS, diagnosed from 1981 to 2014, from the Translational Breast Cancer Research Consortium (TBCRC 038) and the Washington University Repository of Archival Human Breast Tissue (RAHBT). Patients initially diagnosed with DCIS, with either DCIS or invasive recurrence (cases; mean follow up 5.8 years) were m...
2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI)
Ductal Carcinoma In Situ (DCIS) is a non-obligatory precursor of Invasive Breast Cancer. It is th... more Ductal Carcinoma In Situ (DCIS) is a non-obligatory precursor of Invasive Breast Cancer. It is the most common mammographically detected breast cancer. Predicting DCIS progression to invasive ductal carcinoma is a major clinical challenge due to the lack of a uniform classification system in the diagnosis and prognostication of this disease. To characterise the tissue microecology of DCIS, we proposed and tested the model "DCIS-Identification model" based on Generative Adversarial Networks (GAN) for detection and segmentation of DCIS ducts from multiplex immunohistochemistry (IHC) staining samples. We also trained a Spatially Constrained Convolutional Neural Network (SC-CNN) to detect and classify single cells based on their CA9 and FOXP3 expression. The DCIS-Identification model was evaluated on 8 whole slide images, resulting in an average Dice score of 0.95 for the segmentation performance. The single cell identification framework was tested on 10 randomly selected whole slide sections, achieving the average accuracy of 88.6% in a 5 fold cross validation scheme. With the proposed pipeline, we efficiently integrated deep learning, computational pathology and spatial statistics to report distinct differences in the microenvironments of DCIS and IDC/DCIS samples. The proposed pipeline provides a tool for a better understanding of the mechanism of tumours in DCIS and IDC/DCIS cases.
<b>Summary</b> This metadata record provides details of the data supporting the claim... more <b>Summary</b> This metadata record provides details of the data supporting the claims of the related manuscript: "Unmasking the immune microecology of ductal carcinoma in situ with deep learning". The data consist of immunohistochemistry (IHC) haematoxylin and eosin (H&amp;E) staining images of grade 2-3 pure ductal carcinoma in situ (DCIS) and DCIS adjacent to invasive cancer (adjacent DCIS) samples. The related study aimed to characterise tissue spatial architecture and the microenvironment of DCIS via design and validation of a new deep learning pipeline. <b> </b> <b>Data access</b> All training data, including the fully anonymised raw H&amp;E image tiles and pathological annotations as binary marks, as well as Python code, are available in the corresponding author's GitHub: https://github.com/pathdata/HE_Tissue_Segmentation. Requests for data access for the Duke samples can be submitted to E. Shelley Hwang ([email protected])<sup> </sup>and Yinyin Yuan ([email protected]). Data underlying Figures 4 and 6 are in the files "Ext_validData_DCIS_DAVE_Fig4_data.csv" and "Ext_validData_DCIS_DAVE_Fig6_data.csv", included with this metadata record. The images used as representative examples in Figure 8 are listed in the file "Figure 8 image details.xlsx", included with this metadata record. <b>Name of Institutional Review Board or ethics committee that approved the study</b> The study was approved by the institutional review board of Duke with a waiver of the requirement to obtain informed consent.
SUMMARYDuctal carcinoma in situ (DCIS) is the most common precursor of invasive breast cancer (IB... more SUMMARYDuctal carcinoma in situ (DCIS) is the most common precursor of invasive breast cancer (IBC), with variable propensity for progression. We have performed the first multiscale, integrated profiling of DCIS with clinical outcomes by analyzing 677 DCIS samples from 481 patients with 7.1 years median follow-up from the Translational Breast Cancer Research Consortium (TBCRC) 038 study and the Resource of Archival Breast Tissue (RAHBT) cohorts. We identified 812 genes associated with ipsilateral recurrence within 5 years from treatment and developed a classifier that was predictive of DCIS or IBC recurrence in both cohorts. Pathways associated with recurrence include proliferation, immune response, and metabolism. Distinct stromal expression patterns and immune cell compositions were identified. Our multiscale approach employed in situ methods to generate a spatially resolved atlas of breast precancers, where complementary modalities can be directly compared and correlated with con...
Supplementary Figures and Tables. This file contains the following figures and tables: Validation... more Supplementary Figures and Tables. This file contains the following figures and tables: Validation of image analysis with pathological scores. Figure S2. Comparison of Voronoi and Square tessellation. Figure S3. Comparing the Morisita-Horn index with Pearson correlation. Figure S4. Heatmap to show correlation among Pearson correlation and the Morisita-Horn index computed over different scales with square (S) or Voronoi (V) tessellation. Figure S5. Association of immune-cancer correlation with survival according to different spatial scales. Figure S6. Optimizing the Morisita index cutoff in the discovery cohort. Figure S7. Immune-cancer colocalization sub-stratifies clinical parameters and immune abundance, estrogen receptor (ER) status and TP53 mutation status in human epidermal growth factor receptor 2 (HER2)â +â subtype. Figure S8. Robustness of prognostic value of Morisita index in HER2-amplified subtype. Figure S9. Intra-slide score variability. Figure S10. Kaplan-Meier curves of...
In mammography, calcifications are one of the most common signs of breast cancer. Detection of su... more In mammography, calcifications are one of the most common signs of breast cancer. Detection of such lesions is an active area of research for computer-aided diagnosis and machine learning algorithms. Due to limited numbers of positive cases, many supervised detection models suffer from overfitting and fail to generalize. We present a one-class, semi-supervised framework using a deep convolutional autoencoder trained with over 50,000 images from 11,000 negative-only cases. Since the model learned from only normal breast parenchymal features, calcifications produced large signals when comparing the residuals between input and reconstruction output images. As a key advancement, a structural dissimilarity index was used to suppress non-structural noises. Our selected model achieved pixel-based AUROC of 0.959 and AUPRC of 0.676 during validation, where calcification masks were defined in a semi-automated process. Although not trained directly on any cancers, detection performance of calcification lesions on 1,883 testing images (645 malignant and 1238 negative) achieved 75% sensitivity at 2.5 false positives per image. Performance plateaued early when trained with only a fraction of the cases, and greater model complexity or a larger dataset did not improve performance. This study demonstrates the potential of this anomaly detection approach to detect mammographic calcifications in a semi-supervised manner with efficient use of a small number of labeled images, and may facilitate new clinical applications such as computer-aided triage and quality improvement.
Beneficial mutations that arise in an evolving asexual population may compete or interact in ways... more Beneficial mutations that arise in an evolving asexual population may compete or interact in ways that alter the overall rate of adaptation through mechanisms such as clonal or functional interference. The application of multiple selective pressures simultaneously may allow for a greater number of adaptive mutations, increasing the opportunities for competition between selectively advantageous alterations, and thereby reducing the rate of adaptation. We evolved a strain of Saccharomyces cerevisiae that could not produce its own histidine or uracil for ~500 generations under one or three selective pressures: limitation of the concentration of glucose, histidine, and/or uracil in the media. The rate of adaptation was obtained by measuring evolved relative fitness using competition assays. Populations evolved under a single selective pressure showed a statistically significant increase in fitness on those pressures relative to the ancestral strain, but the populations evolved on all th...
Crucial transitions in cancer-including tumor initiation, local expansion, metastasis, and therap... more Crucial transitions in cancer-including tumor initiation, local expansion, metastasis, and therapeutic resistance-involve complex interactions between cells within the dynamic tumor ecosystem. Transformative single-cell genomics technologies and spatial multiplex in situ methods now provide an opportunity to interrogate this complexity at unprecedented resolution. The Human Tumor Atlas Network (HTAN), part of the National Cancer Institute (NCI) Cancer Moonshot Initiative, will establish a clinical, experimental, computational, and organizational framework to generate informative and accessible three-dimensional atlases of cancer transitions for a diverse set of tumor types. This effort complements both ongoing efforts to map healthy organs and previous largescale cancer genomics approaches focused on bulk sequencing at a single point in time. Generating single-cell, multiparametric, longitudinal atlases and integrating them with clinical outcomes should help identify novel predictive biomarkers and features as well as therapeutically relevant cell types, cell states, and cellular interactions across transitions. The resulting tumor atlases should have a profound impact on our understanding of cancer biology and have the potential to improve cancer detection, prevention, and therapeutic discovery for better precision-medicine treatments of cancer patients and those at risk for cancer. Cancer forms and progresses through a series of critical transitions-from pre-malignant to malignant states, from locally contained to metastatic disease, and from treatment-responsive to treatment-resistant tumors (Figure ). Although specifics differ across tumor types and patients, all transitions involve complex dynamic interactions between diverse pre-malignant, malignant, and non-malignant cells (e.g., stroma cells and immune cells), often organized in specific patterns within the tumor
Despite increasing evidence supporting the clinical relevance of tumour infiltrating lymphocytes ... more Despite increasing evidence supporting the clinical relevance of tumour infiltrating lymphocytes (TILs) in invasive breast cancer, TIL spatial distribution pattern surrounding ductal carcinoma in situ (DCIS) and its association with progression is not well understood.To characterize the tissue microecology of DCIS, we designed and tested a new deep learning pipeline, UNMaSk (UNet-IM-Net-SCCNN), for the automated detection and simultaneous segmentation of DCIS ducts. This new method achieved the highest sensitivity and recall over cutting-edge deep learning networks in three patient cohorts, as well as the highest concordance with DCIS identification based on CK5 staining.Following automated DCIS detection, spatial tessellation centred at each DCIS duct created the boundary in which local ecology can be studied. Single cell identification and classification was performed with an existing deep learning method to map the distribution of TILs. In a dataset comprising grade 2-3 pure DCIS...
Understanding intratumor heterogeneity is critical to designing personalized treatments and impro... more Understanding intratumor heterogeneity is critical to designing personalized treatments and improving clinical outcomes of cancers. Such investigations require accurate delineation of the subclonal composition of a tumor, which to date can only be reliably inferred from deep-sequencing data (>300x depth). To enable accurate subclonal discovery in tumors sequenced at standard depths (30-50x), we develop a novel computational method that incorporates an adaptive error model into statistical decomposition of mixed populations, which corrects the mean-variance dependency of sequencing data at the subclonal level. Tested on extensive computer simulations and real-world data, this new method, named model-based adaptive grouping of subclones (MAGOS), consistently outperforms existing methods on minimum sequencing depth, decomposition accuracy and computation efficiency. MAGOS supports subclone analysis using single nucleotide variants and copy number variants from one or more samples of...
Kombucha, a fermented tea beverage with an acidic and effervescent taste, is composed of a multis... more Kombucha, a fermented tea beverage with an acidic and effervescent taste, is composed of a multispecies microbial ecosystem with complex interactions that are characterized by both cooperation and conflict. In kombucha, a complex community of bacteria and yeast initiates the fermentation of a starter tea (usually black or green tea with sugar), producing a biofilm that covers the liquid over several weeks. This happens through several fermentative phases that are characterized by cooperation and competition among the microbes within the kombucha solution. Yeast produce invertase as a public good that enables both yeast and bacteria to metabolize sugars. Bacteria produce a surface biofilm which may act as a public good providing protection from invaders, storage for resources, and greater access to oxygen for microbes embedded within it. The ethanol and acid produced during the fermentative process (by yeast and bacteria, respectively) may also help to protect the system from invasio...
Kombucha is a sweetened tea fermented by bacteria and yeast into a carbonated, acidic drink, prod... more Kombucha is a sweetened tea fermented by bacteria and yeast into a carbonated, acidic drink, producing a surface biofilm pellicle (colloquially called a SCOBY) during the process. Typically, liquid and a biofilm pellicle from a previously fermented culture is used as a starter for new cultures; however, there is no standard protocol for growing kombucha in the laboratory. In order to establish a standard protocol with low variability between replicates, we tested whether we could begin a kombucha culture with only well-mixed liquid stock. We found that viable kombucha cultures can be grown from low percentages of initial inoculum stock liquid, that new pellicles can form from liquid alone (with no ‘starter’ pellicle), and that the variation in the pellicle characteristics is lower when only a liquid starter is used (p = 0.0004). We also found that blending the pellicle before including it significantly reduces the variation among replicates, though the final pellicle was abnormal. W...
Within the cancer microenvironment, the growth and proliferation of cancer cells in the primary s... more Within the cancer microenvironment, the growth and proliferation of cancer cells in the primary site as well as in the metastatic site represent a global biological phenomenon. To understand the growth, proliferation and progression of cancer either by local expansion and/or metastasis, it is important to understand the cancer microenvironment and host response to cancer growth. Melanoma is an excellent model to study the interaction of cancer initiation and growth in relationship to its microenvironment. Social evolution with cooperative cellular groups within an organism is what gives rise to multicellularity in the first place. Cancer cells evolve to exploit their cellular environment. The foundations of multicellular cooperation break down in cancer because those cells that misbehave have an evolutionary advantage over their normally behaving neighbors. It is important to classify evolutionary and ecological aspects of cancer growth, thus, data for cancer growth and outcomes need to be collected to define these parameters so that accurate predictions of how cancer cells may proliferate and metastasize can be developed.
Uploads
Papers by Carlo Maley